Conservation PracticeWildlife

Pilot whale buoy detects calls, but cannot yet predict strandings

Ecological / Economic90 / 10
EcologicalEconomic
Evidence / Opinion70 / 30
EvidenceOpinion
Consensus / Contested85 / 15
ConsensusContested
four shale in body of water under cloudy sky
Illustrative image by Vivek Kumar on Unsplash | Unsplash License
Passive acoustic monitoring

Published 7 November 2026. This story may be updated as it develops.

The story at a glance

An autonomous acoustic buoy tested near Onetahua/Farewell Spit detected pilot whale vocalisations in real time. Project Jonah reports a 98.7% recall rate against calls identified through manual review of recordings. 1

The trial has not shown that the system can predict strandings or reduce whale deaths. On the morning of a mass stranding, its call detections remained below the alert threshold then in use. 1

What you need to know

  • The buoy operated for 158 days and recorded calls from pilot whales and dolphins. Its reported 98.7% recall refers to vocalisations found through manual review, not to the prediction of stranding events. 1
  • On 8 January 2026, the buoy recorded pilot whale vocalisations on the morning of a mass stranding. The calls did not reach the alert threshold that was in place at the time. 1
  • Project Jonah says it adjusted the alert settings after manually reviewing those recordings. The account does not establish whether the changed settings would have provided a useful warning before the stranding. 1
  • The intended system would detect whale calls underwater and send information to shore-based responders. The proposed benefit of an earlier warning remains prospective, rather than a demonstrated outcome of the trial. 12
  • New Zealand’s Department of Conservation says the causes of pilot whale mass strandings are not well understood. It identifies shallow, gently sloping waters as one possible factor, alongside the whales’ social behaviour. 3

Why this matters

The trial shows that a buoy can collect and transmit pilot whale call information, but that capability is not the same as a reliable warning of a stranding. For responders to act on alerts, the relationship between calls and strandings, and the time available to respond, will need to be established. 12

What happens next

Project Jonah says further seasons of monitoring are needed to test whether consistent acoustic patterns occur before strandings and to validate the AI. Until that work is done, the project does not claim the system can predict when whales will strand. 12

What we know

  • The first trial took place near Onetahua/Farewell Spit and ran for 158 days. 1
  • The buoy recorded calls from pilot whales and dolphins, and the project reports a 98.7% recall rate when its detections were checked against manual review. 1
  • The system is designed to detect underwater calls and transmit information to shore-based responders. 2

What we don't know

  • It remains unknown whether pilot whale calls show a consistent pattern before strandings across multiple events and seasons. 1
  • The supplied reporting does not establish how reliably detections and alerts will work amid changing underwater sounds, or whether responders would receive enough lead time to affect outcomes. 12
  • The reported recall figure has not been shown in the supplied material to have been independently evaluated. Its relationship to false alarms or missed stranding events is also unclear. 1

Read the reporting

This WildNews briefing synthesises and summarises reporting from the sources below. WildNews did not produce the original reporting. Learn more about how we use AI

Was this story useful?

Your anonymous feedback helps us make future WildNews briefings more useful.

WILDNEWS WEEKLY DIGEST

One clear conservation briefing, every Friday.

Get the essential stories, context and developments in one clear briefing every Friday.